For patients with rare genetic diseases, the analysis of structural variants (SVs) in addition to smaller indels and single nucleotide variants (SNVs) can improve the diagnostic yield by ~15%. SV calling from whole genome sequence (WGS) data yields on average 16 SVs per case using a specialized SV calling pipeline in clinical use at Rady’s Children Hospital, of varying sizes and gene composition. In a standard diagnostic workflow, a considerable amount of time is spent clinically evaluating each of these SVs for pathogenicity in addition to the SNVs. Gene ranking algorithms provide an efficient strategy for quickly identifying causal candidate genes with SNVs, but to date they have not considered SVs. Fabric GEM is Fabric Genomics’ latest AI gene prioritization algorithm that analyzes not only SNVs but SVs as well, in a single step for rapid identification of causal variants for rare genetic diseases. Here we present the performance of Fabric GEM in over 60 rare whole genome disease cases, provided by our collaborators from the Rady’s Children Hospital. Fabric GEM provides AI-based identification of disease-causing variants and an integrated workflow for identifying and reporting on candidate diagnostic SNVS and SVs, improving the speed and diagnostic yield, and reducing the need for manual interpretation of SVs.